Australia's Proposed Automated Decision-Making Bill: Safeguards and Transparency

What is the proposed bill about?

The Australian government has announced plans to introduce a regulatory framework for automated decision-making, with a focus on safeguards and transparency. This comes after the Department of Industry, Science and Resources published a discussion paper on safe and responsible AI in Australia, which outlined current and international AI regulatory frameworks.

The government has committed to strengthening the way federal agencies such as Centrelink and Services Australia use artificial intelligence, and will also revisit privacy legislation. The new framework will be led by the Attorney-General, Michelle Rowland, as part of broader efforts to regulate AI use by government bodies.

What safeguards and transparency measures are under consideration?

The discussion paper highlighted potential regulatory measures that could be implemented, including evaluating the quality of data used to develop AI systems or requiring AI systems to meet certain transparency standards so that consumers and other stakeholders understand the algorithms and decisions behind them.

These assurances could be developed by industry, government, or both, and could be voluntary or mandated by legislation. They could be managed internally or by third parties in the private or public sector.

In addition, the government plans to introduce a second round of privacy reforms aimed at responsibly strengthening, modernising, and simplifying Australia's personal data protection laws, including further improvements to privacy safeguards.

How might this affect innovation and international AI access?

The discussion paper noted that increased regulation could lead to higher regulatory burden and potentially stifle innovation or prevent Australia from accessing advanced AI technologies developed internationally. However, some Australian regulators are already collaborating on guidelines and sharing information about AI use in specific contexts. For example, the National AI Centre's Responsible AI Network, operated by CSIRO, focuses on research into responsible AI solutions for Australian industry.

The paper also mentioned examples of AI applications that may be subject to such measures, including AI robots used in medical procedures and AI in safety-related car components or autonomous vehicles making real-time decisions.

What else is the government doing?

Alongside the regulatory framework, the government has promised to provide strong protections for Australian creators against the improper use of their works by AI, and has announced plans for regulations on large, energy-intensive data centres.

The proposed framework reflects a broader effort to ensure that AI is developed and used safely and responsibly in Australia, while balancing innovation and protections.